NEW ZEALAND JOURNAL OF PHYSIOTHERAPY | 65 INVITED CLINICAL COMMENTARY Physiotherapy management of complex regional pain syndrome Catherine Pollard BSc (Honours) Physiotherapy, Diploma Orthopaedic Medicine Senior Physiotherapist, The Auckland Regional Pain Service, Auckland District Health Board ABSTRACT Complex Regional Pain Syndrome is a painful debilitating condition characterised by sensory, vasomotor, sudomotor, and trophic changes. Traditionally, physiotherapy treatments have been directed at peripheral symptoms, often with limited efficacy. In light of the growing scientific evidence promoting the major role of the central nervous system in the pathogenesis of Complex Regional Pain Syndrome, there has been a shift towards interventions considered to modulate central processing. A systematic review performed in 2009 aimed to assess the evidence regarding the physiotherapy management of Complex Regional Pain Syndrome. Techniques showing some promise include mirror therapy, Graded Motor Imagery, tactile discrimination training, and exposure therapy. This paper aims to elaborate on the scientific framework for these techniques and explore the current research regarding treatment efficacy. Hopefully, further wide dissemination of these ideas will spark more interest from clinical practitioners and clinicians alike in the quest to more completely understand and manage this complex condition. Pollard C (2013) Physiotherapy management of complex regional pain syndrome New Zealand Journal of Physiotherapy 41(2): 65-72. Key Words: Body perception disturbance, Graded Motor Imagery, mirror therapy, tactile discrimination training, exposure therapy. INTRODUCTION Most physiotherapists either have encountered, or will encounter, a challenging case of Complex Regional Pain Syndrome (CRPS), often, but not necessarily following a patient’s injury, myocardial infarction, or stroke (De Mos 2007, Veldman 1993). Traditionally, physiotherapy treatments have focussed mostly on attempted modification or management of peripheral symptoms, often with limited efficacy. More recently, spurred by scientific advances identifying the significant role of the central nervous system in the pathogenesis of CRPS, techniques which focus on central processes have been developed (Moseley 2010). Treatment strategies including mirror therapy, Graded Motor Imagery, tactile discrimination training and exposure therapy have been explored in one guise or another. However, there is little in the literature as to how these worlds of scientific evidence and best clinical practice come together. This paper addresses and attempts to bridge this divide, reviewing the scientific research that informs our adoption of these novel treatment techniques. Diagnostic Criteria and Pathophysiology The clinical features of CRPS include burning pain, allodynia (pain from a non-painful stimulus) and hyperalgesia (increased response to a painful stimulus); motor disturbances ranging from decreased range, speed, co-ordination of movement, tremor and muscle spasms; changes in vascular tone, temperature, skin colour, sweating and oedema; trophic changes to skin, hair, nails and perceptual disturbances with distortions to the body-self (Harden and Bruehl 2006, Lewis and McCabe 2010). There are two types of CRPS described: CRPS-1 can occur spontaneously or following trauma, with the symptoms unrelated to the region of a single nerve, and disproportionate to the inciting event. CRPS-2 occurs in association with nerve damage (Merskey and Bogduk 1994). The management of these are similar; however, it is important to identify the presence of nerve injury in case further intervention is warranted. The exact cause of CRPS is still not fully understood, however there are a number of proposed pathophysiological mechanisms which contribute to the overall symptoms. Neurogenic inflammation, which involves the amplification of cytokines, bradykinins, endothelin, neuropeptide CGRP and Substance P, has been demonstrated in people who developed CRPS after injury (Birklein and Schmelz 2008, Guo et al 2004). It is postulated the elevation of these inflammatory mediators occurs as a result of inadequate inactivation after their release, so they continue to promote inflammation (Birklein and Kingery 2009). Another suggestion is that more receptors are available to receive these inflammatory mediators (Birklein and Kingery 2009). The overall effect is increased temperature, skin reddening, protein extravasation, oedema and augmented nociceptive stimulation. The role of the sympathetic nervous system (SNS) in CRPS has remained controversial. It was originally proposed that the SNS was the main driver for CRPS symptoms, hence its previous name Reflex Sympathetic Dystrophy. Under normal circumstances sympathetic activity does not impact on the discharge of nociceptors; however in the case of CRPS, nociceptors appear to be under the influence of the SNS. This is referred to as sympathetically maintained pain (Raja et al 2010). In people with CRPS the epidermis of the skin within the region of hyperalgesia has been shown to contain a greater density of the receptors involved in sympathetically maintained pain compared to pain free skin and normal controls (Raja et al 2010). It was agreed however, that the SNS was not the sole cause of CRPS, as sympathetic nerve blocks did not provide significant relief for a number of patients (Galer et al 2001). Based on physiological and functional imaging studies there is substantial evidence that in persistent pain states, reorganisation of the primary somatosensory cortex (the Penfield ‘homunculus’) (Flor 2003, Flor et al 2009), the secondary somatosensory cortex
Physiotherapy management of complex regional pain syndrome
Dec 09, 2022
Welcome message from author
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
INVITED CLINICAL COMMENTARY
Physiotherapy management of complex regional pain syndrome Catherine Pollard BSc (Honours)
Physiotherapy, Diploma Orthopaedic Medicine Senior Physiotherapist, The Auckland Regional Pain Service, Auckland District Health Board
AbstrAct
Complex Regional Pain Syndrome is a painful debilitating condition characterised by sensory, vasomotor, sudomotor, and trophic changes. Traditionally, physiotherapy treatments have been directed at peripheral symptoms, often with limited efficacy. In light of the growing scientific evidence promoting the major role of the central nervous system in the pathogenesis of Complex Regional Pain Syndrome, there has been a shift towards interventions considered to modulate central processing. A systematic review performed in 2009 aimed to assess the evidence regarding the physiotherapy management of Complex Regional Pain Syndrome. Techniques showing some promise include mirror therapy, Graded Motor Imagery, tactile discrimination training, and exposure therapy. This paper aims to elaborate on the scientific framework for these techniques and explore the current research regarding treatment efficacy. Hopefully, further wide dissemination of these ideas will spark more interest from clinical practitioners and clinicians alike in the quest to more completely understand and manage this complex condition.
Pollard C (2013) Physiotherapy management of complex regional pain syndrome New Zealand Journal of Physiotherapy 41(2): 65-72.
Key Words: Body perception disturbance, Graded Motor Imagery, mirror therapy, tactile discrimination training, exposure therapy.
IntroductIon
Most physiotherapists either have encountered, or will encounter, a challenging case of Complex Regional Pain Syndrome (CRPS), often, but not necessarily following a patient’s injury, myocardial infarction, or stroke (De Mos 2007, Veldman 1993). Traditionally, physiotherapy treatments have focussed mostly on attempted modification or management of peripheral symptoms, often with limited efficacy. More recently, spurred by scientific advances identifying the significant role of the central nervous system in the pathogenesis of CRPS, techniques which focus on central processes have been developed (Moseley 2010). Treatment strategies including mirror therapy, Graded Motor Imagery, tactile discrimination training and exposure therapy have been explored in one guise or another. However, there is little in the literature as to how these worlds of scientific evidence and best clinical practice come together. This paper addresses and attempts to bridge this divide, reviewing the scientific research that informs our adoption of these novel treatment techniques.
diagnostic criteria and Pathophysiology
The clinical features of CRPS include burning pain, allodynia (pain from a non-painful stimulus) and hyperalgesia (increased response to a painful stimulus); motor disturbances ranging from decreased range, speed, co-ordination of movement, tremor and muscle spasms; changes in vascular tone, temperature, skin colour, sweating and oedema; trophic changes to skin, hair, nails and perceptual disturbances with distortions to the body-self (Harden and Bruehl 2006, Lewis and McCabe 2010).
There are two types of CRPS described: CRPS-1 can occur spontaneously or following trauma, with the symptoms unrelated to the region of a single nerve, and disproportionate to the inciting event. CRPS-2 occurs in association with nerve damage (Merskey and Bogduk 1994). The management of these are similar; however, it is important to identify
the presence of nerve injury in case further intervention is warranted.
The exact cause of CRPS is still not fully understood, however there are a number of proposed pathophysiological mechanisms which contribute to the overall symptoms. Neurogenic inflammation, which involves the amplification of cytokines, bradykinins, endothelin, neuropeptide CGRP and Substance P, has been demonstrated in people who developed CRPS after injury (Birklein and Schmelz 2008, Guo et al 2004). It is postulated the elevation of these inflammatory mediators occurs as a result of inadequate inactivation after their release, so they continue to promote inflammation (Birklein and Kingery 2009). Another suggestion is that more receptors are available to receive these inflammatory mediators (Birklein and Kingery 2009). The overall effect is increased temperature, skin reddening, protein extravasation, oedema and augmented nociceptive stimulation.
The role of the sympathetic nervous system (SNS) in CRPS has remained controversial. It was originally proposed that the SNS was the main driver for CRPS symptoms, hence its previous name Reflex Sympathetic Dystrophy. Under normal circumstances sympathetic activity does not impact on the discharge of nociceptors; however in the case of CRPS, nociceptors appear to be under the influence of the SNS. This is referred to as sympathetically maintained pain (Raja et al 2010). In people with CRPS the epidermis of the skin within the region of hyperalgesia has been shown to contain a greater density of the receptors involved in sympathetically maintained pain compared to pain free skin and normal controls (Raja et al 2010). It was agreed however, that the SNS was not the sole cause of CRPS, as sympathetic nerve blocks did not provide significant relief for a number of patients (Galer et al 2001).
Based on physiological and functional imaging studies there is substantial evidence that in persistent pain states, reorganisation of the primary somatosensory cortex (the Penfield ‘homunculus’) (Flor 2003, Flor et al 2009), the secondary somatosensory cortex
66 | NEW ZEALAND JOURNAL OF PHYSIOTHERAPY
(Pleger et al 2006), and the motor cortex can occur (Cohen et al 1991). It has been demonstrated that the degree of cortical reorganisation is directly related to the intensity of CRPS pain and the extent of hyperalgesia (Pleger et al 2005).
In the case of CRPS the cortical representation of the affected limb is smaller than that of the unaffected limb, with digit representations moving closer together (Juttonen et al 2002, Maihofner et al 2003, Pleger et al 2004). This can produce affects such as body perception disturbance, whereby people with CRPS describe their limb as feeling abnormal in terms of shape and size (Moseley 2005b), temperature (Lewis et al 2007), position (Lewis et al 2010) and orientation (Schwoebel et al 2001). It can produce feelings so intense that the limb no longer feels like the participant’s own (Lewis et al 2007). It is postulated that this reorganisation can alter cortical processing, instating a conflict between sensory feedback and motor output. It has been shown that inducing a sensorimotor incongruence in normal participants can provoke sensations of spontaneous pain and feelings of peculiarity (McCabe et al 2005) and exacerbate pain in people with fibromyalgia (McCabe et al 2007). It is therefore credible that cortical reorganisation contributes to the pain experienced within CRPS. Cortical reorganisation can also produce motor dysfunction, leading to abnormal movement patterns during reaching and grasping tasks (Maihofner et al 2007).
In summary, these various mechanisms contribute to the multitude of symptoms that can develop in a person with CRPS.
Physiotherapy Management
Over the years many different treatment modalities have been utilised for the management of CRPS, including medical management (analgesics, steroids, supplements) and interventional treatments (sympathetic nerve blocks, sympathectomy, amputation and spinal cord stimulator insertion). It is well recognised however, that physiotherapy plays an important role in the standard treatment of CRPS.
Physiotherapy encompasses a large array of different treatment techniques and modalities. In order to gain a clearer insight into the efficacy of the varied physiotherapy interventions for the treatment of adult CRPS-1 a systematic review of the literature was performed (Daly and Bialocerkowski 2009). An electronic search was conducted for the period 1987-2007 using various databases and searches of textbooks on pain. Each study was appraised by the Australian National Health and Medicine Research Council (NHMRC) hierarchy of evidence and the Critical Review form for Qualitative Studies: 180 articles were found, of which 166 were excluded as they did not meet the inclusion criteria. There were 11 articles included in the systematic review. After analysing and comparing the data regarding the effectiveness of the different treatments, the authors concluded that Graded Motor Imagery (GMI) produced the greatest benefit in terms of reducing pain when compared to conventional physiotherapy and medical management. There was reasonable evidence for modalities such as mirror therapy, desensitisation training, and graded exposure; however, there was no evidence to support the effectiveness of transcutaneous nerve stimulation or stress loading exercise. The study highlighted a distinct lack of high quality research on physiotherapy management of CRPS.
This review aims to expand this systematic review by exploring the latest scientific and clinical based research developments pertaining to these techniques and discuss how they may be applied in a therapeutic setting. In addition, it explores the current research utilising recent modalities such as prism glasses and virtual reality for managing CRPS.
Mirror therapy
Mirror therapy aims to create an illusion of normality in the affected limb. It was introduced by Ramachandran in 1992, for use with phantom limb pain and has since been adapted to aid in the management of numerous conditions, including stroke and pain after wrist fracture and hand surgery (Ramachandran and Alschuler 2009). When used for CRPS, mirror therapy involves concealing the affected limb behind the mirror, while the non-affected limb is positioned so that its reflection is superimposed to where the affected limb should be. The brain has been shown to prioritise visual input over proprioceptive input (Rock and Victor 1964), so when the unaffected limb moves it appears as though the affected limb is functioning normally.
The mechanisms of action for mirror therapy are still not fully understood. There are a number of theories described in the literature including increased attention to the limb, improved ownership of the limb (McCabe 2011), activation of the mirror neurone system (Matthys et al 2009, Rothgangel et al 2011), and a reduction of sensorimotor incongruence (Ramachandran et al 1995).
Mirror therapy has been shown to have positive and negative effects on the symptoms of CRPS (McCabe 2011). It is postulated that the discrepancies in results are due to differing methods of execution. According to McCabe (2011) mirror therapy should be performed with both limbs moving in a bilateral synchronous manner, so the person can feel the movement at the same time as observing the reflection of the normal limb moving. If movement of the affected limb is not performed in synchrony with the observed reflection, conflicting sensory feedback and motor output will be exaggerated and CRPS pain can be increased (McCabe 2011). Acerra and Moseley (2004) demonstrated that pain could be evoked in the affected limb of CRPS participants when the unaffected limb was stimulated in front of a mirror (via light touch, sharp touch and the application of cold). Interestingly, only participants with CRPS experienced pain, it did not occur in participants with similar pain symptoms (but no signs of CRPS-1) or control participants.
Mirror therapy also appears to have differing effects in the acute and chronic phases of CRPS. McCabe et al (2003) performed a pilot study which involved eight participants with CRPS-1 practicing mirror therapy for six weeks. It was demonstrated that visual feedback from the mirror significantly lowered pain intensity in acute CRPS-1 (less than eight weeks). These analgesic effects were prolonged with increasing duration of mirror therapy. In the intermediate stages of the disease (less than one year) mirror therapy reduced stiffness. Unfortunately, there was no beneficial outcome for the three chronic cases. These findings concur with other studies. In acute CRPS, Cacchio et al (2009) demonstrated an improvement to CRPS symptoms, whereas for chronic CRPS Tichelaar et al (2007) reported a poor response to mirror therapy.
NEW ZEALAND JOURNAL OF PHYSIOTHERAPY | 67
When CRPS symptoms persist, patients can experience more physical impairments with changes in muscle strength, contractures, joint stiffness, or motor control. This can place more restrictions on the movement of the affected limb and further increase the incongruence between the affected limb and the mirror image. In these instances it is proposed that mirror therapy may overwhelm the sensitised system therefore exacerbating pain to a greater extent (Moseley 2005a). It has been suggested that a graded approach to cortical activation utilising techniques to activate cortical regions affiliated with movement preparation but not movement execution may be more suitable, as suggested to occur in Graded Motor Imagery (described in more detail in the following section) (Moseley 2005a). This theory was supported when it was demonstrated that during GMI, mirror therapy only imparted an effect when it followed imagery (Moseley 2005a).
In summary, the research indicates that mirror therapy can assist with pain reduction and improve function in the early stages of CRPS. Considering that it is an inexpensive and accessible form of treatment that can be performed within the clinic and continued at the patient’s home, there is a basis for its use in early rehabilitation. In regards to chronic CRPS, there is limited efficacy when used as a first line treatment and in some instance it can exacerbate CRPS symptoms. Caution should be made to ensure patients are instructed on the appropriate technique, to minimise potential side-effects.
Graded Motor Imagery
GMI follows a progressive three-stage motor imagery programme. In stage 1, participants see a series of photographic flash cards, and are asked to identify (as quickly as possible) whether the depiction is of a left or right limb. In stage 2, participants imagine moving the affected limb into the position demonstrated on the photograph, while the affected hand rests comfortably. Stage 3 involves mirror therapy, whereby both limbs are moved to adopt simple postures as demonstrated on the photograph (Mosley 2004).
GMI is considered to exert its effects through sequential activation of distinct (ordered) stages of brain function (Moseley 2005a). Parsons and Fox (1998) used positron-emitting tomography to image brain activation (through blood-flow measures) during right / left judgement tasks (stage 1). A large amount of activity was shown in the pre-motor and supplementary motor regions and the cerebellum, however there was no activity in the primary somatosensory and motor cortices. Imagery (stage 2) has been shown to activate the pre-motor, primary somatosensory and motor cortices (Lotze et al 1999). This indicates that stage 1 activates brain centres involved in higher order aspects of motor control and movement preparation without physical movement of the limb, prior to progressing to stage 2 where activation of the motor cortices occur (Moseley 2005a). This theory was supported during a clinical trial in which 20 participants with chronic CRPS-1 of one hand were randomly allocated to undertake the three components of the GMI programme in different orders (Moseley 2005a). It was demonstrated that participants who followed the sequenced GMI stages (stages 1, 2 then 3) had better outcomes with reduced pain rating and increased functional task ability (measured using the task-specific numeric rating scale) than participants who did not follow the sequence. It also showed that imagined movements were only successful in
producing measurable improvement when they followed hand laterality recognition; and mirror movements were only useful when they followed imagined movements.
Early support for effective utilisation of GMI was demonstrated in a randomised controlled trial involving 13 participants with chronic CRPS-1 following non-complicated wrist fractures (Mosley 2004). Participants were randomly allocated into either a GMI group following the three stage programme or a control group who did not receive treatment. Each stage involved intensive repetition, with exercises practised three times an hour, every waking hour, for two weeks before being progressed to the next stage. On completion of the GMI programme there was a significant reduction in the neuropathic pain scale (by approximately 20 points, on a 100 point scale), an improvement in swelling and reduced limb laterality recognition time. These improvements were maintained for at least six weeks after completion of treatment. The outcome measures for the control group did not change. However, when two of the control participants crossed over to GMI, there was a significant reduction in all three variables.
This study was repeated with a larger sample size including people with phantom limb pain after amputation, brachial plexus avulsion injuries and a more heterogeneous group of CRPS-1 patients. The results showed that pain decreased and function increased for the GMI group relative to the control group; however pain reduction was about 50% less in this study than the previous one (Moseley 2006).
Based on the success of these studies, GMI has been adopted by clinics worldwide. Reports are now being published to discuss the clinical implications of this technique. Johnson et al (2012) performed an audit to assess the outcomes of GMI used within two CRPS speciality centres in the UK. For practical reasons the GMI protocol deviated from that used in the studies by Moseley, with reduced face to face contact, increased duration of the stages, and reduced frequency of practice. Although this makes comparison debateable, it provides a more realistic view of the efficacy of GMI when applied in real-life clinical situations. Unfortunately, the outcomes from this study would suggest that the clinical application of GMI may not be as promising as anticipated. When assessing pain intensity, the participants reported the ‘worst’ pain intensity reduced but the ‘average’ pain intensity remained the same following treatment. On the whole, only 3 out of the 32 patients who started GMI achieved a 50% pain reduction and in 12 out of the 32 patients, pain actually increased with treatment. Lagueux et al (2012) also utilised a modified version of GMI in a clinical trial based on 7 patients with CRPS present for less than 6 months. The results indicated a reduction in pain but no statistically or clinically significant difference to function.
It seems plausible that GMI may provide an avenue to start rehabilitation at a manageable level for a patient who complains that pain is too severe to perform any kind of limb movement. By regressing rehabilitation to a point whereby only the cortical regions involved in movement preparation are activated, pain may be provoked to a lesser extent. This could then be progressed in a steady manner to promote greater cortical activation, prior to commencing functional activation. However, as Johnson et al (2012) identified, there are some cases where pain can be intensified during its use. Further
68 | NEW ZEALAND JOURNAL OF PHYSIOTHERAPY
research to identify potential subgroup populations where GMI may be unsuitable, as well as clearer recommendations for the application of GMI e.g. frequency of practice, duration of stages will assist to optimise the use of GMI in clinical practice.
tactile discrimination
Tactile discrimination is slower in a CRPS-affected limb than in an unaffected limb (Moseley et al 2009) and in some cases, mislocalisation of sensory stimulation is present in the affected limb. Maihöfner et al (2006) demonstrated that when touching the digits of an affected CRPS hand, the sensation was felt to be in another place within the same hand in 8 out of 24 participants tested. It was also noted that the presence of mechanical hyperalgesia was a significant predictor for the incidence of sensory mislocalisation. These occurrences are considered to be related to cortical reorganisation. Flor et al (2001) demonstrated that the extent of reorganisation correlates with the magnitude of pain, and the degree of tactile acuity of the affected region. It has been suggested that tactile information processing is ‘spatially’ related (where the body is in space) rather than somatotopically defined (the body position in accordance to its location within the homunculus). Moseley et al (2009) studied ten participants with CRPS in a single arm. Participants received pairs of vibro-tactile stimuli, one delivered to each hand, at various asynchronies. They were asked to identify which hand had been stimulated first by releasing a foot switch to indicate left or right. This was performed with the arms held each side of the midline and then with the arms crossed over midline. The point at which participants were equally likely to select either hand was compared between conditions and between those with left and right-sided symptoms. The results showed that when arms were not crossed, the participants prioritised stimuli from the unaffected limb over those from the affected limb. In other words, it took participants longer to recognise and/or respond to the stimulus applied to the affected arm. When the arms were crossed the…
Physiotherapy management of complex regional pain syndrome Catherine Pollard BSc (Honours)
Physiotherapy, Diploma Orthopaedic Medicine Senior Physiotherapist, The Auckland Regional Pain Service, Auckland District Health Board
AbstrAct
Complex Regional Pain Syndrome is a painful debilitating condition characterised by sensory, vasomotor, sudomotor, and trophic changes. Traditionally, physiotherapy treatments have been directed at peripheral symptoms, often with limited efficacy. In light of the growing scientific evidence promoting the major role of the central nervous system in the pathogenesis of Complex Regional Pain Syndrome, there has been a shift towards interventions considered to modulate central processing. A systematic review performed in 2009 aimed to assess the evidence regarding the physiotherapy management of Complex Regional Pain Syndrome. Techniques showing some promise include mirror therapy, Graded Motor Imagery, tactile discrimination training, and exposure therapy. This paper aims to elaborate on the scientific framework for these techniques and explore the current research regarding treatment efficacy. Hopefully, further wide dissemination of these ideas will spark more interest from clinical practitioners and clinicians alike in the quest to more completely understand and manage this complex condition.
Pollard C (2013) Physiotherapy management of complex regional pain syndrome New Zealand Journal of Physiotherapy 41(2): 65-72.
Key Words: Body perception disturbance, Graded Motor Imagery, mirror therapy, tactile discrimination training, exposure therapy.
IntroductIon
Most physiotherapists either have encountered, or will encounter, a challenging case of Complex Regional Pain Syndrome (CRPS), often, but not necessarily following a patient’s injury, myocardial infarction, or stroke (De Mos 2007, Veldman 1993). Traditionally, physiotherapy treatments have focussed mostly on attempted modification or management of peripheral symptoms, often with limited efficacy. More recently, spurred by scientific advances identifying the significant role of the central nervous system in the pathogenesis of CRPS, techniques which focus on central processes have been developed (Moseley 2010). Treatment strategies including mirror therapy, Graded Motor Imagery, tactile discrimination training and exposure therapy have been explored in one guise or another. However, there is little in the literature as to how these worlds of scientific evidence and best clinical practice come together. This paper addresses and attempts to bridge this divide, reviewing the scientific research that informs our adoption of these novel treatment techniques.
diagnostic criteria and Pathophysiology
The clinical features of CRPS include burning pain, allodynia (pain from a non-painful stimulus) and hyperalgesia (increased response to a painful stimulus); motor disturbances ranging from decreased range, speed, co-ordination of movement, tremor and muscle spasms; changes in vascular tone, temperature, skin colour, sweating and oedema; trophic changes to skin, hair, nails and perceptual disturbances with distortions to the body-self (Harden and Bruehl 2006, Lewis and McCabe 2010).
There are two types of CRPS described: CRPS-1 can occur spontaneously or following trauma, with the symptoms unrelated to the region of a single nerve, and disproportionate to the inciting event. CRPS-2 occurs in association with nerve damage (Merskey and Bogduk 1994). The management of these are similar; however, it is important to identify
the presence of nerve injury in case further intervention is warranted.
The exact cause of CRPS is still not fully understood, however there are a number of proposed pathophysiological mechanisms which contribute to the overall symptoms. Neurogenic inflammation, which involves the amplification of cytokines, bradykinins, endothelin, neuropeptide CGRP and Substance P, has been demonstrated in people who developed CRPS after injury (Birklein and Schmelz 2008, Guo et al 2004). It is postulated the elevation of these inflammatory mediators occurs as a result of inadequate inactivation after their release, so they continue to promote inflammation (Birklein and Kingery 2009). Another suggestion is that more receptors are available to receive these inflammatory mediators (Birklein and Kingery 2009). The overall effect is increased temperature, skin reddening, protein extravasation, oedema and augmented nociceptive stimulation.
The role of the sympathetic nervous system (SNS) in CRPS has remained controversial. It was originally proposed that the SNS was the main driver for CRPS symptoms, hence its previous name Reflex Sympathetic Dystrophy. Under normal circumstances sympathetic activity does not impact on the discharge of nociceptors; however in the case of CRPS, nociceptors appear to be under the influence of the SNS. This is referred to as sympathetically maintained pain (Raja et al 2010). In people with CRPS the epidermis of the skin within the region of hyperalgesia has been shown to contain a greater density of the receptors involved in sympathetically maintained pain compared to pain free skin and normal controls (Raja et al 2010). It was agreed however, that the SNS was not the sole cause of CRPS, as sympathetic nerve blocks did not provide significant relief for a number of patients (Galer et al 2001).
Based on physiological and functional imaging studies there is substantial evidence that in persistent pain states, reorganisation of the primary somatosensory cortex (the Penfield ‘homunculus’) (Flor 2003, Flor et al 2009), the secondary somatosensory cortex
66 | NEW ZEALAND JOURNAL OF PHYSIOTHERAPY
(Pleger et al 2006), and the motor cortex can occur (Cohen et al 1991). It has been demonstrated that the degree of cortical reorganisation is directly related to the intensity of CRPS pain and the extent of hyperalgesia (Pleger et al 2005).
In the case of CRPS the cortical representation of the affected limb is smaller than that of the unaffected limb, with digit representations moving closer together (Juttonen et al 2002, Maihofner et al 2003, Pleger et al 2004). This can produce affects such as body perception disturbance, whereby people with CRPS describe their limb as feeling abnormal in terms of shape and size (Moseley 2005b), temperature (Lewis et al 2007), position (Lewis et al 2010) and orientation (Schwoebel et al 2001). It can produce feelings so intense that the limb no longer feels like the participant’s own (Lewis et al 2007). It is postulated that this reorganisation can alter cortical processing, instating a conflict between sensory feedback and motor output. It has been shown that inducing a sensorimotor incongruence in normal participants can provoke sensations of spontaneous pain and feelings of peculiarity (McCabe et al 2005) and exacerbate pain in people with fibromyalgia (McCabe et al 2007). It is therefore credible that cortical reorganisation contributes to the pain experienced within CRPS. Cortical reorganisation can also produce motor dysfunction, leading to abnormal movement patterns during reaching and grasping tasks (Maihofner et al 2007).
In summary, these various mechanisms contribute to the multitude of symptoms that can develop in a person with CRPS.
Physiotherapy Management
Over the years many different treatment modalities have been utilised for the management of CRPS, including medical management (analgesics, steroids, supplements) and interventional treatments (sympathetic nerve blocks, sympathectomy, amputation and spinal cord stimulator insertion). It is well recognised however, that physiotherapy plays an important role in the standard treatment of CRPS.
Physiotherapy encompasses a large array of different treatment techniques and modalities. In order to gain a clearer insight into the efficacy of the varied physiotherapy interventions for the treatment of adult CRPS-1 a systematic review of the literature was performed (Daly and Bialocerkowski 2009). An electronic search was conducted for the period 1987-2007 using various databases and searches of textbooks on pain. Each study was appraised by the Australian National Health and Medicine Research Council (NHMRC) hierarchy of evidence and the Critical Review form for Qualitative Studies: 180 articles were found, of which 166 were excluded as they did not meet the inclusion criteria. There were 11 articles included in the systematic review. After analysing and comparing the data regarding the effectiveness of the different treatments, the authors concluded that Graded Motor Imagery (GMI) produced the greatest benefit in terms of reducing pain when compared to conventional physiotherapy and medical management. There was reasonable evidence for modalities such as mirror therapy, desensitisation training, and graded exposure; however, there was no evidence to support the effectiveness of transcutaneous nerve stimulation or stress loading exercise. The study highlighted a distinct lack of high quality research on physiotherapy management of CRPS.
This review aims to expand this systematic review by exploring the latest scientific and clinical based research developments pertaining to these techniques and discuss how they may be applied in a therapeutic setting. In addition, it explores the current research utilising recent modalities such as prism glasses and virtual reality for managing CRPS.
Mirror therapy
Mirror therapy aims to create an illusion of normality in the affected limb. It was introduced by Ramachandran in 1992, for use with phantom limb pain and has since been adapted to aid in the management of numerous conditions, including stroke and pain after wrist fracture and hand surgery (Ramachandran and Alschuler 2009). When used for CRPS, mirror therapy involves concealing the affected limb behind the mirror, while the non-affected limb is positioned so that its reflection is superimposed to where the affected limb should be. The brain has been shown to prioritise visual input over proprioceptive input (Rock and Victor 1964), so when the unaffected limb moves it appears as though the affected limb is functioning normally.
The mechanisms of action for mirror therapy are still not fully understood. There are a number of theories described in the literature including increased attention to the limb, improved ownership of the limb (McCabe 2011), activation of the mirror neurone system (Matthys et al 2009, Rothgangel et al 2011), and a reduction of sensorimotor incongruence (Ramachandran et al 1995).
Mirror therapy has been shown to have positive and negative effects on the symptoms of CRPS (McCabe 2011). It is postulated that the discrepancies in results are due to differing methods of execution. According to McCabe (2011) mirror therapy should be performed with both limbs moving in a bilateral synchronous manner, so the person can feel the movement at the same time as observing the reflection of the normal limb moving. If movement of the affected limb is not performed in synchrony with the observed reflection, conflicting sensory feedback and motor output will be exaggerated and CRPS pain can be increased (McCabe 2011). Acerra and Moseley (2004) demonstrated that pain could be evoked in the affected limb of CRPS participants when the unaffected limb was stimulated in front of a mirror (via light touch, sharp touch and the application of cold). Interestingly, only participants with CRPS experienced pain, it did not occur in participants with similar pain symptoms (but no signs of CRPS-1) or control participants.
Mirror therapy also appears to have differing effects in the acute and chronic phases of CRPS. McCabe et al (2003) performed a pilot study which involved eight participants with CRPS-1 practicing mirror therapy for six weeks. It was demonstrated that visual feedback from the mirror significantly lowered pain intensity in acute CRPS-1 (less than eight weeks). These analgesic effects were prolonged with increasing duration of mirror therapy. In the intermediate stages of the disease (less than one year) mirror therapy reduced stiffness. Unfortunately, there was no beneficial outcome for the three chronic cases. These findings concur with other studies. In acute CRPS, Cacchio et al (2009) demonstrated an improvement to CRPS symptoms, whereas for chronic CRPS Tichelaar et al (2007) reported a poor response to mirror therapy.
NEW ZEALAND JOURNAL OF PHYSIOTHERAPY | 67
When CRPS symptoms persist, patients can experience more physical impairments with changes in muscle strength, contractures, joint stiffness, or motor control. This can place more restrictions on the movement of the affected limb and further increase the incongruence between the affected limb and the mirror image. In these instances it is proposed that mirror therapy may overwhelm the sensitised system therefore exacerbating pain to a greater extent (Moseley 2005a). It has been suggested that a graded approach to cortical activation utilising techniques to activate cortical regions affiliated with movement preparation but not movement execution may be more suitable, as suggested to occur in Graded Motor Imagery (described in more detail in the following section) (Moseley 2005a). This theory was supported when it was demonstrated that during GMI, mirror therapy only imparted an effect when it followed imagery (Moseley 2005a).
In summary, the research indicates that mirror therapy can assist with pain reduction and improve function in the early stages of CRPS. Considering that it is an inexpensive and accessible form of treatment that can be performed within the clinic and continued at the patient’s home, there is a basis for its use in early rehabilitation. In regards to chronic CRPS, there is limited efficacy when used as a first line treatment and in some instance it can exacerbate CRPS symptoms. Caution should be made to ensure patients are instructed on the appropriate technique, to minimise potential side-effects.
Graded Motor Imagery
GMI follows a progressive three-stage motor imagery programme. In stage 1, participants see a series of photographic flash cards, and are asked to identify (as quickly as possible) whether the depiction is of a left or right limb. In stage 2, participants imagine moving the affected limb into the position demonstrated on the photograph, while the affected hand rests comfortably. Stage 3 involves mirror therapy, whereby both limbs are moved to adopt simple postures as demonstrated on the photograph (Mosley 2004).
GMI is considered to exert its effects through sequential activation of distinct (ordered) stages of brain function (Moseley 2005a). Parsons and Fox (1998) used positron-emitting tomography to image brain activation (through blood-flow measures) during right / left judgement tasks (stage 1). A large amount of activity was shown in the pre-motor and supplementary motor regions and the cerebellum, however there was no activity in the primary somatosensory and motor cortices. Imagery (stage 2) has been shown to activate the pre-motor, primary somatosensory and motor cortices (Lotze et al 1999). This indicates that stage 1 activates brain centres involved in higher order aspects of motor control and movement preparation without physical movement of the limb, prior to progressing to stage 2 where activation of the motor cortices occur (Moseley 2005a). This theory was supported during a clinical trial in which 20 participants with chronic CRPS-1 of one hand were randomly allocated to undertake the three components of the GMI programme in different orders (Moseley 2005a). It was demonstrated that participants who followed the sequenced GMI stages (stages 1, 2 then 3) had better outcomes with reduced pain rating and increased functional task ability (measured using the task-specific numeric rating scale) than participants who did not follow the sequence. It also showed that imagined movements were only successful in
producing measurable improvement when they followed hand laterality recognition; and mirror movements were only useful when they followed imagined movements.
Early support for effective utilisation of GMI was demonstrated in a randomised controlled trial involving 13 participants with chronic CRPS-1 following non-complicated wrist fractures (Mosley 2004). Participants were randomly allocated into either a GMI group following the three stage programme or a control group who did not receive treatment. Each stage involved intensive repetition, with exercises practised three times an hour, every waking hour, for two weeks before being progressed to the next stage. On completion of the GMI programme there was a significant reduction in the neuropathic pain scale (by approximately 20 points, on a 100 point scale), an improvement in swelling and reduced limb laterality recognition time. These improvements were maintained for at least six weeks after completion of treatment. The outcome measures for the control group did not change. However, when two of the control participants crossed over to GMI, there was a significant reduction in all three variables.
This study was repeated with a larger sample size including people with phantom limb pain after amputation, brachial plexus avulsion injuries and a more heterogeneous group of CRPS-1 patients. The results showed that pain decreased and function increased for the GMI group relative to the control group; however pain reduction was about 50% less in this study than the previous one (Moseley 2006).
Based on the success of these studies, GMI has been adopted by clinics worldwide. Reports are now being published to discuss the clinical implications of this technique. Johnson et al (2012) performed an audit to assess the outcomes of GMI used within two CRPS speciality centres in the UK. For practical reasons the GMI protocol deviated from that used in the studies by Moseley, with reduced face to face contact, increased duration of the stages, and reduced frequency of practice. Although this makes comparison debateable, it provides a more realistic view of the efficacy of GMI when applied in real-life clinical situations. Unfortunately, the outcomes from this study would suggest that the clinical application of GMI may not be as promising as anticipated. When assessing pain intensity, the participants reported the ‘worst’ pain intensity reduced but the ‘average’ pain intensity remained the same following treatment. On the whole, only 3 out of the 32 patients who started GMI achieved a 50% pain reduction and in 12 out of the 32 patients, pain actually increased with treatment. Lagueux et al (2012) also utilised a modified version of GMI in a clinical trial based on 7 patients with CRPS present for less than 6 months. The results indicated a reduction in pain but no statistically or clinically significant difference to function.
It seems plausible that GMI may provide an avenue to start rehabilitation at a manageable level for a patient who complains that pain is too severe to perform any kind of limb movement. By regressing rehabilitation to a point whereby only the cortical regions involved in movement preparation are activated, pain may be provoked to a lesser extent. This could then be progressed in a steady manner to promote greater cortical activation, prior to commencing functional activation. However, as Johnson et al (2012) identified, there are some cases where pain can be intensified during its use. Further
68 | NEW ZEALAND JOURNAL OF PHYSIOTHERAPY
research to identify potential subgroup populations where GMI may be unsuitable, as well as clearer recommendations for the application of GMI e.g. frequency of practice, duration of stages will assist to optimise the use of GMI in clinical practice.
tactile discrimination
Tactile discrimination is slower in a CRPS-affected limb than in an unaffected limb (Moseley et al 2009) and in some cases, mislocalisation of sensory stimulation is present in the affected limb. Maihöfner et al (2006) demonstrated that when touching the digits of an affected CRPS hand, the sensation was felt to be in another place within the same hand in 8 out of 24 participants tested. It was also noted that the presence of mechanical hyperalgesia was a significant predictor for the incidence of sensory mislocalisation. These occurrences are considered to be related to cortical reorganisation. Flor et al (2001) demonstrated that the extent of reorganisation correlates with the magnitude of pain, and the degree of tactile acuity of the affected region. It has been suggested that tactile information processing is ‘spatially’ related (where the body is in space) rather than somatotopically defined (the body position in accordance to its location within the homunculus). Moseley et al (2009) studied ten participants with CRPS in a single arm. Participants received pairs of vibro-tactile stimuli, one delivered to each hand, at various asynchronies. They were asked to identify which hand had been stimulated first by releasing a foot switch to indicate left or right. This was performed with the arms held each side of the midline and then with the arms crossed over midline. The point at which participants were equally likely to select either hand was compared between conditions and between those with left and right-sided symptoms. The results showed that when arms were not crossed, the participants prioritised stimuli from the unaffected limb over those from the affected limb. In other words, it took participants longer to recognise and/or respond to the stimulus applied to the affected arm. When the arms were crossed the…
Related Documents
